Am 03.04.2012 um 11:21 schrieb Ursula Winkler:

> Reuti wrote:
>> Hi,
>> 
>> Am 03.04.2012 um 10:48 schrieb Ursula Winkler:
>> 
>>  
>>> Dear members of gridengine,
>>> 
>>> I have problems running mpi jobs with SGE. The error message is:
>>>                                          "unable to find pe_start file 
>>> "/softw".
>>> 
>>> Unfortunately not the whole path is printed so I can just guess.
>>> 
>>> I created with qmon --> Parallel Environment Configuration the Environment: 
>>> "mpi" -->
>>> Start Proc Args: /software/sge/mpi/startmpi.sh -catch_rsh $pe_hostfile
>>> Stop Proc Args: /software/sge/mpi/stopmpi.sh
>>>    
>> 
>> which MPI implementation do you intend to use? With Open MPI and MPICH2 you 
>> get a Tight Integration nowadays without and special procedures.
>>  
> MVAPICH2. How do I set up the parallel environment without the "special 
> procedures"?
>> Nevertheless ist should work.
>> 
>> Can you please print the value of $2 in start_proc_args and make a `stat` of 
>> that file/path in question - the error is written by start_proc_args?
>>  
> The error is written by "qstat -j <Job-Nr.>" and in the Spool-Logfile of the 
> Master. What do you mean with $2 in start_proc_args?

The start_proc_args defined script get called with the parameters you specify 
there. Before any processing in the script is done, the supplied parameters 
will be set as local environment variables:

$1 should be the string "-catch_rsh"
$2 should be the expanded string of $pe_hostfile, so we can check what file is 
used later on.

Which MVAPICH scripts do you use? The default supplied in the $SGE_ROOT/mpi was 
for MPICH1 only IIRC.

-- Reuti
_______________________________________________
users mailing list
[email protected]
https://gridengine.org/mailman/listinfo/users

Reply via email to